Multi-robot Task Assignment Based On Group Intelligence

In recent years,with the development of modern science and technology,the performance of mobile robots has been continuously improved,especially in communication technology,electronic communication technology,and control cooperation technology.Multi-robot systems are used in intelligent design,military fields,industrial fields,education fields,and entertainment.The fields have been applied to varying degrees.The application of mobile robots in the field of multi-tasking is becoming more and more common.However,with the frequent emergence of complex and difficult tasks in actual environments,how to build an effective cooperative relationship between multi-robot systems has become a key issue in task allocation.Therefore,this paper proposes two algorithms for the efficient allocation of cooperative tasks(that is,tightly coupled tasks)and non-cooperative tasks(that is,loose tasks).First,in order to solve the loose task assignment,this paper proposes a task assignment algorithm based on improved threshold.After years of research,it has been proved that the traditional threshold method has good dynamic distribution.However,its disadvantages are that it cannot handle complex tasks and the threshold is difficult to set,which affects learning efficiency and allocation efficiency.In order to improve the shortcomings of the traditional threshold method,this paper proposes a task assignment algorithm based on the improved threshold.The traditional threshold method is combined with the task matching evaluation model to realize the learning process of task assignment.The threshold can be converged to a reasonable value through continuous iterative execution The value thus causes each robot to perform a relatively matching task.Secondly,for tightly coupled task assignment,this paper proposes a multi-robot task assignment algorithm based on an improved ant colony algorithm.First,put the ants on the task as evenly as possible in groups,and initially set up a reference set for each task by arranging and combining.Then,the ant colony initially explores the robot randomly through the pheromone and adds it to the alliance solution.Medium,until the solution can complete the task,record this alliance solution.All ants in the group are assigned for tightly coupled tasks.Record the ant colony algorithm corresponding to the most adaptive alliance solution for each task in this round of iteration,and compare the solution with a better reference set to update the reference set for each task.Elements and pheromones on tasks and robot paths until the end condition is met.This algorithm is combined with the improved threshold method to deal with mixed tasks in practical problems.Finally,this paper validates the rationality and effectiveness of these two methods through experiments.By improving the traditional threshold method,the optimization ratio of loose tasks is effectively improved,and the improved ant colony algorithm is used to improve the convergence performance and tasks of the alliance solution.Distribution efficiency.